Key Points

  • The Federal Circuit affirmed that post-priority experimental data showing claimed variants do not work is admissible to prove non-enablement, distinguishing cases involving later-developed technology.
  • Pioneer's own experiments showed only 2 of 8 tested enzymes exhibited the claimed dual herbicide-degrading function despite meeting the patent's structural requirements.
  • The claimed genus encompassed approximately 2.4 times 10 to the 106th power enzyme variants, while the specification disclosed only two working examples.
  • The court applied the Supreme Court's Amgen v. Sanofi standard that the more one claims, the more one must enable.
  • The decision was nonprecedential but clarifies the line between excluding later-developed technology and admitting post-priority evidence about then-makeable species.

The Federal Circuit on Friday affirmed a Patent Trial and Appeal Board decision invalidating Pioneer Hi-Bred International Inc.'s herbicide-tolerance enzyme patent, holding that the company's own experimental data—conducted a decade after the patent's priority date—properly supported a finding that the claims were not enabled.

Judge Prost wrote for a unanimous panel that included Judges Dyk and Cunningham. The nonprecedential decision rejected Pioneer's argument that the Board erred by considering evidence generated years after the patent application was filed.

At issue was U.S. Patent No. 11,371,055, which describes enzymes that make crops resistant to two classes of herbicides: pyridyloxyacetate auxins such as triclopyr and phenoxy auxins such as 2,4-D. Inari Agriculture Inc. challenged the patent in a post-grant review proceeding, arguing that the claimed genus of enzymes was "unfathomably large."

The Board agreed with Inari. The patent's claims covered enzymes "having at least 85% sequence identity" with a disclosed sequence and containing a particular structural motif. That definition, the Board found, encompassed approximately 2.4 times 10 to the 106th power possible enzyme variants—a number greater than the estimated atoms in the universe. The patent's specification, meanwhile, disclosed only two working examples.

Pioneer did not dispute the size of the genus. Instead, it argued that the 85% sequence identity threshold was a "common quality running throughout the claimed genus" that ensured more than 90% of variants would exhibit the dual herbicide-degrading function. Pioneer also pointed to Figure 2 of the patent, which purported to identify residues that could be varied without affecting function.

But even applying that guidance, the genus would still include more than 10 to the 66th power possible variants, the Board found.

What proved most damaging to Pioneer's case was its own experimental data, submitted after the priority date. Pioneer tested eight enzymes that met the claimed structural requirements—all with at least 85% sequence identity. Only two exhibited the dual herbicide-degrading function, and weakly, with "only 1% and 2% . . . triclopyr activity."

Pioneer argued on appeal that it was legal error for the Board to rely on post-priority data because "[e]nablement must be determined . . . as of the priority date." The company cited In re Hogan and In re Entresto, cases that barred the use of later-developed technology to challenge enablement.

The Federal Circuit disagreed, finding that Pioneer's reliance on those cases was misplaced. Judge Prost explained that In re Hogan rejected an examiner's use of "new technology developed after the priority date," where an amorphous polymer that did not exist when the application was filed later fell within the claim scope. That, the court said, would "impose an impossible burden on inventors."

This case was different. "[A]ll tested variants of the claimed genus could have been created at the time of the priority date, although they were tested by Pioneer afterward," Judge Prost wrote. Citing the court's earlier decision in Amgen Inc. v. Sanofi, she wrote that "[t]he use of post-priority-date evidence to show that a patent does not disclose a representative number of species of a claimed genus is proper."

The court noted that Pioneer appeared to have forfeited the post-priority data argument by not raising it before the Board, but proceeded to reject it on the merits regardless.

The panel also declined to disturb the Board's decision to credit Inari's expert, Dr. Silverstone, over Pioneer's objection that the expert lacked sufficient enzymology experience. The Board found that Pioneer had not explained "what difference this additional direct [enzymology] experience would have provided," and the Federal Circuit saw no abuse of discretion.

On the core enablement question, the court invoked the Supreme Court's reasoning in Amgen Inc. v. Sanofi: "[T]he more one claims, the more one must enable." The court found that "[t]he massive disparity between the amount claimed and the amount disclosed supports the Board's finding of lack of enablement."

The court noted that Pioneer's own data contradicted its expert's predictions. Two variants that differed by only six amino acids—none in positions Figure 2 identified as important—produced opposite results, with one exhibiting the claimed function and the other not. Substantial evidence, the court concluded, supported the Board's finding that the specification "fail[ed] to provide guideposts that would have illuminated a path toward embodiments at the 85% sequence identity level."

The decision affirms the Board's final written decision in PGR2023-00022.